The notion of recording lessons learned during a project’s progress and preserving them for posterity (post deployment) seems to have achieved nearly universal consensus as being good. Yet I’ve been frustrated enough (and indifferent enough with respect to my career) to state that, typically, the only lesson learned is that lessons are never learned [1].
Lessons learned can, of course, be construed as a subset of retrospective, or post-mortem, analysis: the subject domain of this interesting, 30-page paper (which has 46 references). Section 2.1, “Software Project Retrospectives,” should be mandatory reading for all managers involved in non-trivial software projects. The paper chiefly addresses the impact of a second obstacle, namely, development teams’ geographical distributions, which sometimes span many time zones. The authors present their cloud-based ARCA tool, RCA standing for root cause analysis, and some initial empirical results from comparison with 35 existing RCA tools. These results are clearly, methodically, and thoroughly presented via perspicuous narrative and detailed tables and diagrams.
The authors consider “the most frequently lacking feature” of current RCA tools to be the synchronization (“syncing”) mechanism for simultaneous co-creation of cause-effect diagrams. At a higher level of abstraction are the three dimensions of group support systems (GSSs): communication support; process structuring; and information processing. The paper substantiates the ARCA tool’s complementing of the retrospective method in realizing GSSs’ three dimensions.
I think that the usefulness of this outstanding paper far transcends its special focus on the problems of widely distributed development teams. All software managers should study it carefully.